Sixteen tests designed to measure aspects of fluid intelligence are to be analyzed along with personality and other ability measures using data obtained from samples of 8th grade children, one subsample (N equals 209) from poor minority-group neighborhoods and one subsample (N equals 624) from well-off majority-group neighborhoods. Scales for the experimental tests are to be developed in ways that will tend to maximize the internal consistency reliabilities and discriminabilities within the sample of minority-group children but will also provide satisfactory reliability and discriminability within the sample of majority-group children. The interrelationships among the experimental variables and the other variables, and the differences between the samples, are to be described using widely understood methods of mathematical and statistical analysis.